What is situated evolution?

Abstract

Abstract-In this paper we discuss the notion of situated evolution. Our treatment includes positioning situated evolution on the map of evolutionary processes in terms of time-and space-embeddedness, and the identification of decentralization as an orthogonal property. We proceed with a selected overview of related literature in the categories of our interest. This overview enables us to distill further detailes that distinguish the encountered methods. As it turns out the essential differences can be captured through the mechanics of selection and fertilization. These insights are aggregated into a new model called the Situated Evolution Method, which is then used to provide a fine-grained map of exisiting work. I. BACKGROUND AND OBJECTIVESThe background of this paper is a research project 1 concerned with a group of robots that operate in a challenging environment and permanently adapt their controllers in order to increase their task performance. Evolution is chosen as the principal method of adaptation, hence evolutionary computing (EC) is expected to supply the technical machinery to enable successful adaptation on-the-fly. This choice draws our attention to evolutionary algorithms, expecting much existing work that can be used to drive the evolutionary mechanics in a group of evolving robots.Looking around in EC soon reveals that there is a large variety of evolutionary algorithms, such as Evolutionary Programming, Evolution Strategies, Genetic Algorithms, Genetic Programming, Differential Evolution, Particle Swarm Optimization, and more. These 'dialects' show a great diversity in algorithmic details, but still fit under a unified model of evolutionary algorithms and are mainly used to solve optimization and machine learning problems [7], [4]. In general, evolutionary algorithms are typically applied as problem solvers pursuing an optimal solution in a search space defined by the problem at hand.However, the collective robotics application we envision, has a very different look-and-feel than a genetic algorithm solving the travelling salesman problem. Intuitively, it is a different kind of evolutionary system, where the essence is not optimization in an abstract search space, but a population of reproducing agents that undergoes selection in a (physical or virtual) environment. For the time being, we will use the term situated evolution for such systems. Such systems have been studied within robotics, artificial life, adaptive multiagent research etc., and by the different perspective and the corresponding terminology it is not clear how do they relate to mainstream evolutionary computing.The inspiration for the present paper lies in the experience that on the one hand, systems based on situated evolution {mc.schut,e.haasdijk,ae.eiben}@few.vu.nl are clearly evolutionary, but on the other hand are very different from mainstream evolutionary algorithms. This forms an intellectual challenge, as we would like to understand the nature of these differences. Additionally, understanding the (dis)similarities has a...

Introduction to Evolutionary Computing

“…In other words, such systems feature 'situated evolution' or 'physically embodied evolution' as discussed in [371]. (The first edition of this book treated it very briefly in Sect.…”

In Vivo Veritas: Towards the Evolution of Things

“…This is certainly not the case in natural evolution. In general, situated evolution without central orchestration will rely on local, non-synchronized decisions regarding birth and death [36]. Hence, the existing individuals can be removed without adding new ones and new ones can be added without discarding old ones first.…”

It’s Fate: A Self-organising Evolutionary Algorithm

“…In situated evolution in general, and in evolutionary swarm robotics in particular, a single oracle has a number of drawbacks [11]. Firstly, it forms a single point of failure, secondly, it may limit scalability as it may not be able to process all the information about the individuals in a timely manner, and thirdly, it may not be reachable for certain individuals if the distance exceeds the feasible (or cost effective) range of communication.…”